There is a well-known need for an implantable glucose sensor in the treatment of diabetes. We have a sensor that shows promise. A simple version of the sensor has been implanted intravenously in dogs and has functioned for more than l00 days without the need for recalibration, and has clear potential for operation over longer periods. We now plan to develop a similar sensor for long-term implantation in tissues. The sensor design will be optimized for tissue application by an iterative modeling and prototyping approach, with which we have extensive experience. Key criteria have been identified for design optimization for this application, including: glucose range, oxygen sensitivity and availability, transient response to glucose and oxygen, sensor size and shape, signal magnitude, and others. Several promising sensor designs will be simulated, compared based on the above criteria, fabricated as prototypes having the essential components necessary to validate simulation results, and tested in vitro. The endpoints of this Phase I effort will be: (l) an optimized and thoroughly simulated sensor design, and (2) a prototype sensor, tested in vitro, incorporating the most promising design features. The complete sensor unit, including electronics, implant housing, and biocompatible interface will be manufactured and tested in vivo in Phase II. PROPOSED COMMERCIAL APPLICATIONS: The management of diabetes consumes 15% of total national health care expenditures, 25% of MediCare expenditures, and is responsible for an enormous, untold cost in human resources. These costs could be reduced substantially by better blood glucose control resulting from improved insulin replacement, made possible by continuous monitoring of blood glucose. Several million people with diabetes could benefit.